*Manuscript Click here to view linked References Wing resonances in a dead-leaf-mimic katydid 1

1 Wing resonances in a new dead-leaf-mimic katydid 1 2 3 2 (: Pterochrozinae) from the Andean cloud forests 4 5 6 3 7 8 1 1 2 1 9 4 ANDREW BAKER , FABIO A. SARRIA-S , GLENN K. MORRIS , THORIN JONSSON 10 11 5 AND FERNANDO MONTEALEGRE-Z1* 12 13 14 6 15 16 7 1School of Life Sciences, Joseph Banks Laboratories, Green Lane, Lincoln, LN6 7DL, UK 17 18 8 2Department of Biological Sciences, University of Toronto Mississauga, 3359 Mississauga 19 20 21 9 Road North, Mississauga, Ontario, Canada, L5L 1C6 22 23 10 24 25 26 11 Running title: Wing resonances in a leaf-mimic katydid 27 28 12 29 30 31 13 * Corresponding author. E-mail: [email protected] 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 2

14 Abstract 1 2 3 15 Day-camouflaged leaf-mimic katydids Typophyllum spp. have a remarkable way of evading 4 5 16 predators as male and female forewings appear as bite-damaged leaves complete with necrotic 6 7 17 spots. As in all other katydids, males produce sound signals to attract females by rubbing their 8 9 10 18 forewings together. The biophysical properties of these special leaf-like forewings remain 11 12 19 obscure. Here we study the wing mechanics and resonances of Typophyllum spurioculis, a 13 14 15 20 new of leaf-mimic katydid with a broad distribution in the Andes from Western 16 17 21 Ecuador to the middle Central Cordillera in Colombia. This species performs an unusual 18 19 20 22 laterally directed aposematic display, showing orange spots that simulate eyes at the leg base. 21 22 23 At night, males are conspicuous by their loud, audible calling songs, which exhibit two 23 24 25 24 spectral peaks at ca. 7 and 12 kHz. Using micro-scanning laser Doppler vibrometry we find 26 27 25 the effective sound radiators of the wings (speculae) vibrate with three modes of vibration, 28 29 26 two of which include the frequencies observed in the calling song. Remarkably, this 30 31 32 27 resonance is preserved in the parts of the wings mimicking necrotic leaves, which are in 33 34 28 theory not specialised for sound production. The eyespot function is discussed. 35 36 37 29 38 39 30 Keywords: Katydid, Bush-, Mimetism, Stridulation, Resonance 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 3

31 Introduction 1 2 3 32 Katydid species of the Typophyllum are masters of camouflage and mimicry. Both 4 5 33 sexes exhibit remarkable forewing adaptations, resembling chewed leaves living and dead, 6 7 34 with necrotic spots, marginal feeding ‘bites’ or even skeletonized ocellata (Braun, 2015b; 8 9 10 35 Nickle and Castner, 1995). During the daytime, males and females rest on trees and bushes 11 12 36 overlooked by diurnal predators. Individuals adopt postures making them almost 13 14 15 37 indistinguishable from the surrounding foliage: the head is applied to a plant branch and the 16 17 38 ’s long antennae projected forward and down along the axis of the same branch (Fig. 1). 18 19 20 39 All known species are nocturnally active, when the males produce loud calling songs to 21 22 40 attract females. (Braun, 2015a; Braun, 2015b; Montealegre-Z and Morris, 1999). Females are 23 24 25 41 larger than males and their tegmina are not only comparatively larger, they also differ in 26 27 42 venation pattern and shape from the male wings (Braun 2002; Nickle, 1992). 28 29 43 During a collecting expedition in 1996 in Ucumarí (Colombian Andean cloud forests, see 30 31 32 44 methods), GKM and FM-Z discovered a male of a curious Typophyllum species, which they 33 34 45 later briefly mentioned (without description) (Morris and Montealegre-Z, 2001). In the same 35 36 37 46 article they also figure an analysis of its calling song: a spectrum with two maximum 38 39 47 harmonically-related peaks at around 7 and 13 kHz. More than 20 years later, FS, AB, and 40 41 42 48 FMZ discovered the same species in the montane forest of Santa Lucía, a cloud forest reserve 43 44 49 located in the Pichincha province of Ecuador. The species is very abundant in this locality, 45 46 50 with males dominating the audible acoustic space in the nights. 47 48 49 51 Male katydids produce calling songs to attract potential mates by rubbing their forewings, a 50 51 52 mechanism known as stridulation (Gwynne, 2001; Robinson and Hall, 2002). A sharp lobe on 52 53 54 53 the right wing (the scraper) is swept upon a vein lined with precisely arranged cuticular teeth 55 56 54 on the left wing. A large specular wing cell (adjacent to the scraper) on the right wing, the so- 57 58 59 55 called mirror, subsequently couples the low amplitude vibrations, produced by this initial file- 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 4

56 scraper interaction (Bailey and Broughton, 1970; Broughton, 1964; Morris, 1999; Morris and 1 2 57 Pipher, 1967; Robinson, 1990) to the air. 3 4 5 58 A particular characteristic of the acoustic signals in the genus Typophyllum is that the calls of 6 7 59 the males exhibit a very sharp spectrum with high Q and that most species known so far 8 9 10 60 communicate at about 20 kHz (Braun 2002; Braun, 2015a; Braun, 2015b; Montealegre-Z and 11 12 61 Morris, 1999; Morris et al., 1989) (Q=quality factor, which describes the properties of a 13 14 62 damped resonator or oscillatory system (Bennet-Clark, 1999b). Would the presence of two 15 16 17 63 harmonically related peaks in the call of our new species Typophyllum spurioculis (here 18 19 64 described) involve two independent resonators in the wings, or could not the same speculum 20 21 22 65 or another wing area vibrate to produce harmonics? 23 24 66 In katydids, the stridulatory fields of left and right tegmina are asymmetric. The stridulatory 25 26 27 67 area of the left forewing where the file resides, is usually damped to sound vibration, while 28 29 68 that of the right wing is heavily involved in sound radiation (Chivers et al., 2017; 30 31 69 Montealegre-Z and Postles, 2010; Sarria-S et al., 2016; Sarria-S et al., 2014). In contrast to 32 33 34 70 katydids, in crickets and grigs (with morphologically symmetric or nearly symmetric wings) 35 36 71 both left and right wings contribute more or less equally to sound radiation (Chivers et al., 37 38 39 72 2016). In theory, the observed wing asymmetry in katydids will affect the radiating sound 40 41 73 field, having a negative effect on the output power. In other words, a sound radiator involving 42 43 44 74 symmetric wings is analogous to having two synchronised speakers. Therefore, katydids, 45 46 75 having only one functional sound radiator, sacrifice one speaker for sound radiation. Katydids 47 48 have resolved this problem by developing additional morphological adaptations to amplify 49 76 50 51 77 their signal (Bennet-Clark, 1998). For example, special wing positions that favour signal 52 53 78 broadcasting (Stumpner et al., 2013), wing inflations (Hemp et al., 2013), and exaggerated 54 55 56 79 pronotal extensions that work as resonators (Jonsson et al., 2017; Morris and Mason, 1995). 57 58 80 However the wings of male Typophyllum, seem not to have special adaptations for resonance 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 5

81 (Braun, 2015a; Montealegre-Z and Mason, 2005; Morris et al., 1989). While the effective 1 2 82 stridulatory area in the right wing, the mirror, is only a small oval portion of the entire large 3 4 5 83 leaf-like wing (Braun 2002; Braun, 2015a; Montealegre-Z and Morris, 1999), the calls of the 6 7 84 male are very loud, suggesting that these might use a combination of resonance and 8 9 10 85 subalar space to enhance the call output. The biophysical properties of the wing structure used 11 12 86 for camouflage and not involved in sound radiation remains elusive. 13 14 87 In this paper we use micro-scanning laser Doppler vibrometry to investigate the wing 15 16 17 88 resonances associated with the production of the species’ acoustic signals, and the resonances 18 19 89 associated with the wing regions involved in camouflage. We also describe a new species of 20 21 22 90 Typophyllum, and document the variation of wing morphology and venation across 23 24 91 individuals. 25 26 27 92 28 29 93 Materials and methods 30 31 32 94 Field sampling 33 34 95 Fieldwork was conducted in Colombia, at Parque Regional Natural Ucumarí (5º N, 76º W). 35 36 37 96 This is located in the eastern slope of the Central Cordillera, above the tiny hamlet El Cedral, 38 39 97 22 km east from Pereira in the Department of Risaralda. A narrow glaciated valley rises into 40 41 42 98 the montane rainforest over elevations of 1850-2600 m, and high waterfalls enter along the 43 44 99 sides. Rainfall is 2000-4000 mm/year and the average temperature 12-18 ºC. Insects were 45 46 100 collected in 1996 within a few kilometres from el Refugio Turistico La Pastora, some 6 km 47 48 49 101 from El Cedral. 50 51 102 We also conducted fieldwork in 2015 in Ecuador, at Santa Lucía cloud forest reserve, located 52 53 54 103 in the heart of the cloud forest on a mountain peak at 1900 metres. Santa Lucía conserves over 55 56 104 1800 acres of montane cloud forest in the Chocó Andean Bioregion, a biodiversity 'hotspot'. 57 58 59 105 Santa Lucía is located in the northwest of the Pichincha province, 80 km north-west of Quito 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 6

106 (00° 07′ 05.9398″ N, 078° 36′ 37.1403″ W). Specimens were localized by their calling song, 1 2 107 mostly from trees and bushes along the main trail a few kilometres from the lodge. 3 4 5 108 6 7 109 Depositories 8 9 10 110 MEUV = Museo de Entomología, Universidad del Valle, Cali, Colombia. 11 12 111 MEUCE = Museo de Entomología, Pontificia Universidad Católica del Ecuador, Quito, 13 14 112 Ecuador. 15 16 17 113 18 19 114 Morphological measurements 20 21 22 115 Body measurements were obtained with a digital calliper (Fowler, Newton, Massachusetts, 23 24 116 USA). All measurements presented are in mm, following the measuring protocols used by 25 26 27 117 Montealegre-Z and Morris (1999). Digital photographs of preserved stridulatory files were 28 29 118 taken on a scanning electron microscope (Inspect S50, FEI, Eindhoven, Holland). 30 31 119 Measurements of inter-tooth spacing were obtained using CorelDraw X4 (Corel Inc. 2005) 32 33 34 120 using the appropriate dimension tool, for details see Montealegre-Z & Mason (Montealegre-Z 35 36 121 and Mason, 2005). 37 38 39 122 40 41 123 Song recordings 42 43 124 Field recordings: In Santa Lucía, Ecuador, four specimens were recorded in the field using a 44 45 46 125 Tascam DR-5 V2, at a sampling rate of 96 kHz. The recorder was placed at 30 cm dorsal to 47 48 126 the singing insect. Insects were captured and taken to the lab facility at the lounge, placed in 49 50 51 127 cylindrical cages of metallic mesh (15 cm high x 8 cm diameter), and recorded again using 52 53 128 the same recording device, positioned at 30 cm from the dorsum of the singing specimen. 54 55 129 56 57 58 130 Lab recordings (Ecuadorean males): In the lab, recordings were performed in a sound- 59 60 131 attenuated booth at the University of Lincoln, at a temperature of 19 °C. The specimens were 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 7

132 placed in metallic cages at 10 cm (dorsal aspect) from a G.R.A.S type 40DD 1/8” condenser 1 2 133 microphone (G.R.A.S. Sound and Vibration, Holte, Denmark). The microphone was 3 4 5 134 connected to a GRAS type 12AA preamplifier, which was, in turned, connected to a 6 7 135 soundboard (USB-6259, National Instruments, Austin, TX), and then to the controlling 8 9 10 136 computer. The microphone was calibrated at 94 dB SPL (re 20 µPa), using a Brüel & Kjaer 11 12 137 sound level calibrator (Type 4231, Brüel & Kjaer, Nærum, Denmark). Data was stored on a 13 14 138 computer hard disk at a sampling rate of 512 kHz. Sound was analysed using Matlab 15 16 17 139 (R2015a, The MathWorks, Inc., Natick, MA, USA). 18 19 140 20 21 22 141 Lab recordings (Colombian male): Laboratory analysis employed Brüel & Kjaer equipment. 23 24 142 A 1/4" (4135) or 1/8" (4138) condenser microphone was connected to a 2606 measuring 25 26 27 143 amplifier or to a 2204 B&K sound level meter. Output from either of these amplifiers was 28 29 144 recorded on a Racal instrumentation tape recorder running at 30" /s (Racal Electronics plc, 30 31 145 Weybridge, UK). Subsequently, the signals were digitized using either a Keithley DAS50 32 33 34 146 digitizing board (Tektronix U.K. Ltd, Bracknell, Berkshire, UK) or Tucker Davis system II 35 36 147 (Tucker-Davis Technologies, Alachua, FL, USA) and then analysed with DADISP software 37 38 39 148 (DSP Development Corporation, Newton, MA, USA). Energy in spectra was only considered 40 41 149 significant if it was no more than 20 dB below the most intense peak frequency. Readings 42 43 44 150 were taken on Fast or Impulse/Hold, usually with a distance of 10 cm from the microphone 45 46 151 tip to the dorsum of the singer. The long axis of the microphone was normal to the 47 48 longitudinal axis of the insect and the microphone cover was always on. Temperatures were 49 152 50 51 153 taken with an Omega HH23 digital thermometer (Omega, Northbank, Manchester). 52 53 154 54 55 56 155 Forewing resonance 57 58 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 8

156 Wing resonance was measured in five male specimens using micro-scanning laser Doppler 1 2 157 vibrometry (LDV; PSV-500, Polytec GmbH, Waldbronn, Germany). For the experiments, the 3 4 5 158 insect was mounted on a brass platform following the procedure described by Sarria-S et al. 6 7 159 (2016). The wings were laterally extended by fixing the axillary sclerites with a 50% mixture 8 9 10 160 of beeswax (Fisher Scientific, Loughborough, UK) and colophonium (Sigma-Aldrich 11 12 161 Company Ltd., Dorset, UK). This procedure only tests the natural resonance of the wings in 13 14 162 response to sound; it eliminates the effect of the subalar space, and leverage effects of scraper 15 16 17 163 motion upon right tegmen radiators during effective stridulation. The stridulatory area of both 18 19 164 tegmina were excited via sympathetic sound stimulation, and scanned with the LDV using 20 21 22 165 800 grid points. 23 24 166 In addition, the entire tegminal surface was also scanned to explore potential resonances 25 26 27 167 outside the stridulatory field. Acoustic stimulation consisted of broadband periodic chirps in 28 29 168 the range of 2-100 kHz. The spectrum of the stimulus was corrected to be flat (± 1.5 dB) at 60 30 31 169 dB SPL at all frequencies. The acoustic signals were generated by the PSV-500 internal data 32 33 34 170 acquisition board (PCI-4451; National Instruments, Austin, TX, USA), amplified (A-400, 35 36 171 Pioneer, Kawasaki, Japan) and passed to a loudspeaker (Ultrasonic Dynamic Speaker Vifa, 37 38 39 172 Avisoft Bioacoustics, Glienicke, Germany) positioned 30 cm from the specimen. The 40 41 173 reference signal was recorded using a 1/8” condenser microphone positioned horizontally at a 42 43 44 174 distance of 2-3 mm from the wings (Brüel & Kjaer, 4138-A-015 and preamplifier model 45 46 175 2670, Brüel & Kjaer, Nærum, Denmark). Laser and sound recordings were all obtained in a 47 48 sound-attenuated booth (1.8 x 1.8 x 2 m) at approximately mid-day in the ’ light 49 176 50 51 177 cycle). Laser experiments were done in five males collected in 2015. 52 53 178 54 55 56 179 Analysis 57 58 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 9

180 For laser measurements of wing vibrations we use FFT (Fast Fourier Transform) with a 1 2 181 rectangular window, resulting in a frequency spectrum for the laser and the microphone signal 3 4 5 182 with a resolution of 7.5 Hz. The laser and microphone signals were then used to calculate the 6 7 183 gain and phase responses. We integrated the results from all the points scanned across the 8 9 10 184 wing surfaces (see details above), oscillation profiles and animations of membrane deflections 11 12 185 were generated for the frequencies of interest. For the final analysis we calculated the transfer 13 14 186 function (H1) of the membrane displacement D (nm) to reference sound level (Pa), to produce 15 16 17 187 the amplitude gain and the phase response of the system at different frequencies. Therefore, 18 19 188 ; 20 21 22 23 189 where, Gab(f) is the cross-spectrum of the displacement signal and reference signal, and Gaa(f) 24 25 190 is the auto-spectrum of the reference signal. 26 27 28 191 29 30 192 Results 31 32 33 193 34 35 36 194 Typophyllum spurioculis sp. nov. 37 38 195 See Figs. 1-3 39 40 196 41 42 43 197 Holotype: MEUCE 1♂, Ecuador, Pichincha, Nanegal, Reserva Santa Lucía. Collector: F. 44 45 198 Montealegre-Z., July 14-20, 2017. Allotype: MEUCE 1♀, Ecuador, Pichincha, Nanegal, 46 47 48 199 Reserva Santa Lucía. Collector: C. Soulsbury, F. Montealegre-Z., July 14-20, 2017 (see 49 50 200 supplementary figure Fig. S1) 51 52 53 201 54 55 202 Paratypes: MEUV 1♂, Colombia, Risaralda, PRN Ucumarí, Collectors: G. K. Morris & F. 56 57 203 Montealegre-Z., 25 May, 1996. MEUCE 3 ♂♂ Ecuador, Pichincha, Nanegal, Reserva Santa 58 59 60 204 Lucía. Collectors: F. Sarria-S & F. Montealegre-Z., Sep, 2014. MEUV 2 ♂♂ Ecuador, 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 10

205 Pichincha, Nanegal, Reserva Santa Lucía. Collector: F. Montealegre-Z., July 1-12, 2015. 1 2 206 MEUCE 1♂, Ecuador, Pichincha, Nanegal, Reserva Santa Lucía. Collector: F. Montealegre- 3 4 5 207 Z., July 1-12, 2015. MEUCE 1♀, Ecuador, Pichincha, Nanegal, Reserva Santa Lucía. 6 7 208 Collector: F. Montealegre-Z., July 1-12, 2015. 8 9 10 209 11 12 210 Etymology: The species is named Typophyllum spurioculis in token of the pairs of vivid 13 14 211 orange maculae on the coxotrochanteral area of pro- and mesothoracic legs on both sides (Fig. 15 16 17 212 1, inset). ‘Spurioculis’ can be translated as ‘with/of false eyes’, i.e., the ‘leaf-mimic with false 18 19 213 eyes’ or the 'False-eyed Typophyllum'. It is a combination of spurius, meaning false in Latin 20 21 22 214 and oculus, meaning eyes in Latin. 23 24 215 25 26 27 216 Diagnosis: This species exhibits an orange vivid coloration in the coxal membranes of the 28 29 217 fore and middle legs. The extension of femur and trochanter relative to the coxa exposes this 30 31 218 membrane (crescent-shaped) giving the appearance of a pair of "eyes" (Fig. 1A, yellow 32 33 34 219 arrows in inset). In all individuals a small black, blunt protrusion lies on the right and left 35 36 220 anterior corners of the coxal spine (Fig. 1A, white arrows in inset). 37 38 39 221 40 41 222 Description: Head: Frons extense, considerably flat, occupying half of the head area as 42 43 44 223 viewed frontal. The other half of the head dominated by clypeus and labrum. Pronotum: 45 46 224 Pronotal disk campaniform, granulose, with small tubercles abundant in the lateral portions. 47 48 Posterior margin of the disk bilobulated, about 2x wider than anterior margin; lobes clearly 49 225 50 51 226 depicted by a medial notch. Disk with a large tubercle in each corner of the distal margin, and 52 53 227 two more tubercles (usually smaller than the corner ones) subequally distributed on the sides 54 55 56 228 of the notch. 57 58 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 11

229 Legs: Fore femora with four spines, forward facing and varying in size (Fig. 2C). Fore coxal 1 2 230 spine blunt, in the form of a black large tubercle (Fig. 1A, inset). Smaller spines located 3 4 5 231 proximal, larger spines distal, the two distal spines sometimes emerging from a broad cuticle 6 7 232 protruding flap. The spines in males are usually very small except for the forelegs. Generally, 8 9 10 233 forelegs and midlegs have 4 and hindlegs have 7-8 spines on the femur. Relatively evenly 11 12 234 spaced, backwards facing spines on the dorsal interior edge of the femora. 13 14 235 Wings: Tegmina of the female (measured from basal to distal) about 1.5x larger than those of 15 16 17 236 the males (Fig. 1A, Table 1). Adult females’ wing marks vary from necrotic spots in the 18 19 237 wings (Fig. 1) to white large marks that resemble fungal patches (see supplementary figure 20 21 22 238 Fig. S1). These are usually absent in males. Venation patterns in males are variable (Fig. 2A), 23 24 239 but a common venation pattern becomes evident across specimens when the wings of several 25 26 27 240 specimens are overlaid (Fig. 2B). Stridulatory file ca. 5 mm long, bearing 160-170 teeth (n=5 28 29 241 males, Table 1, Fig. 3). The file shows a peculiar tooth distribution forming 8-10 clear groups 30 31 242 of teeth with high density (or shorter inter-tooth distances) alternating with 8-10 groups of 32 33 34 243 teeth with low density (larger inter-tooth distances, Fig. 3AB). This file morphology is neatly 35 36 244 associated with the size (length) of individual teeth. Groups with high density of teeth bear 37 38 39 245 larger teeth (Fig. 3B). 40 41 246 Abdomen: Small short pale spines on side of female abdomen. Pale patch with dark border on 42 43 44 247 side of abdomen is common on several specimens, but not in all. Small medial lappets on 45 46 248 abdominal tergites (probably vestigial moss camouflage from earlier nymph stages). 47 48 Genitalia: Male subgenital plate small, distally with a shallow sinusoidal notch. Female 49 249 50 51 250 subgenital plate with a broad v-shaped notch, the depth of the notch is comparable to a 1/3 of 52 53 251 the plate length. Ovipositor up-curved, dorsal valve with 25-30 serrated distal teeth, ventral 54 55 56 252 valve with 18-22 serrated distal teeth (see supplementary figure Fig. S2). These ovipositor 57 58 253 serrations do not occur in nymphs (see supplementary figure Fig. S3). 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 12

254 Variation: The right tegmen of the male has a translucent mirror. 1 2 255 Small black spots all over body of the female (only a few of the males exhibit this). Spots and 3 4 5 256 holes (transparent parts imitating holes on a leaf) are also highly variable. Small light or dark 6 7 257 brown spots tend to appear in-between tegmen veins. A middle cell in the radial field of the 8 9 10 258 tegmina is commonly affected by a spot or 'hole'. 11 12 259 Colouration: Highly variable. Most individuals green, others brown/yellow, necrotic patches 13 14 260 highly variable. One individual had an orange/brown head and body along with green legs, 15 16 17 261 antenna and tegmina. Some tegmina are brown with light or dark brown patches, others are 18 19 262 green with light brown patches. It is unknown if these patches appear with age. Male hind 20 21 22 263 wings are translucent white with green veins. The tips of the hind wings are tinted green. Legs 23 24 264 are usually green but can be brown or a mixture of both (brown individuals can have small 25 26 27 265 yellow patches). Antennae are usually various shades of green, but can also be brown with 28 29 266 faint green patches. Dark line from the base of the antenna to the lower mandible, barely 30 31 267 visible in some brown males. 32 33 34 268 35 36 269 Measurements: see Table 1. 37 38 39 270 40 41 271 Remarks: There are 34 species described in the genus Typophyllum (Cigliano et al., 2017). 42 43 272 We described T. spurioculis as a species new to science based on morphological and acoustic 44 45 46 273 evidence; however we suspect this species might be part of a species complex (within 47 48 274 Typophyllum). We have observed other cloud forest species, which also exhibit the coxal 49 50 51 275 orange eye-spots, and present variable morphology and different calls. But these are so far 52 53 276 observations of a few random specimens for which we do not have the acoustic, and 54 55 277 mechanical evidence presented here. The eye-spots are also present in pre-adult nymphs. 56 57 58 278 However, we do not have data for earlier nymphal stages. 59 60 279 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 13

280 Bioacoustics 1 2 281 The description of the song is based on the recording of the call of a single male specimen 3 4 5 282 collected in Ucumarí Colombia, and five males from Santa Lucía, Ecuador. 6 7 283 Colombian specimen: One male was recorded indoors at a temperature of 23 °C, rather high 8 9 10 284 relative to common field temperatures. This male produced three bouts of sibilant four-pulse 11 12 285 calls, each comprised of two syllables. In T. spurioculis, each syllable consists of two pulses, 13 14 286 one produced during the opening stroke of the wing and the other during the closing (Fig. 15 16 17 287 4A). Here, syllables are therefore equivalent to what other authors term phonatomes, and a 18 19 288 call of several phonatomes is in turn equivalent to a chirp or verse (compare e.g. Jatho et al., 20 21 22 289 1994; Stumpner and Meyer, 2001; Walker and Dew, 1972). Five to six calls were given in 3-6 23 24 290 s with bouts separated by silence of 17-18 seconds. Mean call duration for this male was 95 25 26 27 291 ms (n=13). There is a consistent pause before the concluding pulse, which usually begins 28 29 292 more abruptly. Mean values for the duration of the four high-amplitude pulses in two calls 30 31 293 were 23.3, 17.6, 12.3, and 34.3 ms respectively. The waveform (Fig. 4C) shows the 32 33 34 294 characteristic pattern of two harmonically-related subequal carrier frequencies. The call is 35 36 295 easily heard by a human ear, being almost entirely in the audio range (Fig 4E). There are two 37 38 39 296 harmonically related spectral peaks of comparable intensity. The more intense is near 12.1 40 41 297 kHz while its potential fundamental, 6.8 kHz, is slightly suppressed (mean values for n=13 42 43 44 298 calls, FFT calculated over the whole of each song). 45 46 299 Ecuadorean specimens: five males from Ecuador were recorded at 17 °C under field 47 48 conditions using a portable recorder. From this group, two males were also recorded in 49 300 50 51 301 laboratory conditions at 19 °C in Santa Lucía, Ecuador using the same device. Males of this 52 53 302 population also produced groups of calls, each group consisting of four or five calls (each call 54 55 56 303 containing two syllables), and groups are separated by silent intervals of 18.2 ± 4.2 s (Mean ± 57 58 304 SD, n=5, average measured from six random silent intervals of one recording of each male, 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 14

305 for a total of 30 calls, Fig. 5A). From field recordings, specimens delivered 3 to 4 calls in 10 1 2 306 seconds; a similar pattern was maintained under lab conditions only when more than two 3 4 5 307 males interacted acoustically. Isolated singing males produced a more sporadic pattern, 6 7 308 sometimes producing one call per minute. The average duration of one call (two syllables) of 8 9 10 309 the Ecuadorian males was 117.6 ± 7.1 ms (n = 5, 7 calls randomly selected from each male, 11 12 310 and averaged, Figs. 4B, 5C). The mean pulse duration of the high amplitude pulse was 31.6 ± 13 14 311 3.6 ms (Fig. 4B, 5C). The waveform of the pulse also shows similar characteristics to that of 15 16 17 312 the Colombian male, with two dominant frequencies harmonically related (Fig. 4D). The 18 19 313 major syllable pulse contains 160-170 oscillations. As measured in the Colombian specimen, 20 21 22 314 the call of the Ecuadorean males also shows two consistent spectral peaks of comparable 23 24 315 intensity that are harmonically related. The most intense peak was measured at 13.9 ± 2.2 kHz 25 26 27 316 (n=5, 1 random call per male) and its fundamental occurs at 7.0 ± 1.2 kHz (Fig. 4F). Analysis 28 29 317 of frequency in time shows an unusual frequency modulation (FM) pattern of the two pulses 30 31 318 in each syllable. The FM occurs periodically between 6.3 and 7 kHz, and this is more 32 33 34 319 pronounced in the first syllable of the group (Fig. 5). 35 36 320 Sound Pressure Level (SPL) measured across males in lab conditions was 98.2 ± 2.9 dB (re 37 38 39 321 20 µPa; root mean square over recorded songs with calibrated microphone at 10 cm dorsal 40 41 322 from the singing male). 42 43 44 323 45 46 324 Wing resonances 47 48 Five male specimens were used for LDV experiments. Both LW and RW were stimulated 49 325 50 51 326 with broadband sound and wing resonances obtained by scanning the wing surface using 52 53 327 LDV. These recordings show that only the right mirror plays a major role in sound radiation, 54 55 56 328 while the equivalent left sound-radiating field is highly damped to vibrations (Fig. 6A-F). 57 58 329 Within individuals the mirror shows three modes of vibration at around 6.8 ± 0.7 kHz, 13.5 ± 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 15

330 1.5 kHz, and 16.6 ± 1.9 (n=5 individual averages of 120-130 scanned points in each mirror). 1 2 331 Fig. 6G shows average FFT spectrum across the five males. The observed resonances of 11.3- 3 4 5 332 16.6 kHz are within the range of carrier frequencies measured across specimens. 6 7 333 During the same experiments we also studied the resonances of the entire wings using 8 9 10 334 broadband stimulation (Fig. 7). We divided the wings in broad regions, represented by 11 12 335 numbers in Fig. 7A. Region 1 in the middle of the costal field, region 2 radial/middle field, 13 14 336 region 3 wing apex, and region 4, the stridulatory field. Although regions 1-3 vibrate with 15 16 17 337 lower amplitude than the right stridulatory field, both wings show components of the 18 19 338 resonances observed in the mirror of region 4 (~7, and 10 kHz, ranges observed across five 20 21 22 339 individuals), as well as those frequencies observed in the calling song in these ranges. The 23 24 340 resonant peaks are variable within this range across individuals, but all of them show a 25 26 27 341 common resonant peak between at 10 and 11 kHz, as illustrated in Fig. 7E-H. 28 29 342 30 31 32 343 Discussion 33 34 344 Acoustics and wing resonances 35 36 37 345 Males of most Typophyllum spp. reported so far produce signals approaching acoustic purity 38 39 346 at around 20 kHz (Braun, 2015a; Montealegre-Z and Morris, 1999; Morris et al., 1989). 40 41 42 347 Typophyllum spurioculis is singing lower than most Typophyllum spp. This cloud forest 43 44 348 species became distinctive for the loud, audible call, easy to localise by humans. In most 45 46 349 acoustic the calling song carrier frequency depends on, and closely matches, the 47 48 49 350 resonance of the sound generator (the right mirror and associated cells) (Bennet-Clark, 1999a; 50 51 351 Chivers et al., 2017; Montealegre-Z et al., 2011; Montealegre-Z and Postles, 2010). This is 52 53 54 352 because the tooth strike rate matches the resonance of the sound generator. For example, if the 55 56 353 resonant frequency of the radiating areas is 5 kHz, and the stridulum is excited at a rate of 57 58 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 16

354 5000 teeth per second the system is said to reach resonance and vibrate at maximum 1 2 355 amplitude (Bennet-Clark, 1999a, 2003b; Fletcher, 1992, 2007). 3 4 5 356 We showed here that the mirror vibrates with three modes, two of them within the frequencies 6 7 357 observed in the calling song (~7 and 14 kHz, Fig. 6). These resonances were also observed in 8 9 10 358 other wing regions not directly involved in stridulation (Fig. 7). The two observed peaks 11 12 359 result from one sine wave superimposed on another, which is twice as fast and it gives a 13 14 360 sinusoid with a recurring bump (Fig. 4CD). This phenomenon has been observed in other 15 16 17 361 katydids. Morris, 1980 first realized this in rhinoceros (see Fig. 3 in that paper), 18 19 362 but it has been shown that other species show more than one peak of subequal amplitudes 20 21 22 363 (e.g. Panacanthus cuspidatus in Montealegre-Z and Morris, 2004 and Uchuca halticos in 23 24 364 Montealegre-Z and Morris, 2003). This is what we observe in Typophyllum spurioculis (Fig. 25 26 27 365 4EF). 28 29 366 Our findings suggest that the two harmonically-related peaks observed in the calling song at 30 31 367 around 7 and 14 kHz result from stimulation of any of the two modes, as both resonances are 32 33 34 368 present in the various parts of the wings. Since we do not have mechanical evidence of the 35 36 369 wing motion during sound production we cannot assert which of the two modes is stimulated 37 38 39 370 during the interaction of scraper and file. However, based on the duration of the major pulse 40 41 371 in the syllable (~32 ms), and assuming that this pulse was produced by sweeping the entire 42 43 44 372 file length (~5.3 mm, with some 160-170 teeth), an average velocity of file sweeping could be 45 46 373 inferred to be ~165 mm/s. This velocity and the average inter-tooth distances (0.03 mm) could 47 48 be used to predict the time used by the scraper to jump between two teeth, which will in 49 374 50 51 375 theory correspond to the period (p) of the fundamental oscillation, in other words p=0.03/165, 52 53 376 whose reciprocal gives an instantaneous frequency (f=1/p)  5.5 kHz. This analysis suggests 54 55 56 377 that the wings are stimulated with a tooth strike rate that closely matches the fundamental 57 58 378 frequency of the call (ca 6k teeth/sec, average in Fig. 6). That excitation may cause the low 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 17

379 frequency components across both wings, including the mirror) to release sounds at 6-7 kHz 1 2 380 (see Fig. 7), but also excites the higher vibration mode of the mirror at ca ~12-14 kHz (Fig. 3 4 5 381 6). Proof of this is that the number of file teeth matches the number of fundamental 6 7 382 oscillations in the major pulse of the syllable (see call analysis and file description). In 8 9 10 383 summary, the observed energy peaks in the spectrum of the calling song (Fig. 4EF) are the 11 12 384 result of excitation of the first mode of vibration of the wings, and indirect excitation of the 13 14 385 dominant mode of vibration of the mirror (Fig 6). This form of excitation is not uncommon in 15 16 17 386 Ensifera, and it seems to be a strategy for energy conservation in high-frequency Eneopterinae 18 19 387 crickets (Robillard et al., 2013) 20 21 22 388 At the fundamental frequency, we also observed that both pulses in a syllable are sinusoidally 23 24 389 frequency modulated (Fig. 5C, lower part). This seems to be the result of a default mechanical 25 26 27 390 modulation already encoded in the file morphology. The tooth density (and in turn the 28 29 391 produced frequency produced at constant speed) periodically alternates between small groups 30 31 392 of high-density and small groups of low density of teeth (Fig. 3). We attribute the observed 32 33 34 393 modulation of the pulses in the call to such tooth arrangement. The sweeping of the scraper 35 36 394 over a group of teeth with high density will produce high frequencies, while low frequencies 37 38 39 395 are produced in the segments with low-density of teeth. This is a case of mechanical 40 41 396 frequency modulation, which was first documented in other leaf-mimic katydids singing at 42 43 44 397 higher frequencies (Braun, 2015b; Montealegre-Z, 2005), but not studied in detail so far. FM 45 46 398 has been a widely reported phenomenon observed in the call of insects, mostly in crickets and 47 48 katydids. It was originally reported by Leroy, 1966 in crickets, and subsequently followed by 49 399 50 51 400 measurements of Morris and Pipher, 1967 in the katydid nigropleurum. In 52 53 401 crickets and katydids, FM seems to be the result of the gradual decrease in wing speed during 54 55 56 402 the closing stroke (Montealegre-Z, 2005; Montealegre-Z et al., 2011; Montealegre-Z and 57 58 403 Mason, 2005). However, in field crickets, the envelope of the modulation is a ‘finger print’ of 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 18

404 each male individual (Montealegre-Z et al., 2011). This seems to be associated to the levels of 1 2 405 asymmetry between left and right tegmina (Bennet-Clark, 2003a). Very little is known about 3 4 5 406 the function of FM in the calling song of many species and the only existing neuro- 6 7 407 ethological study in crickets suggests that females prefer non-modulated calls over modulated 8 9 10 408 ones (Hirtenlehner et al., 2013). 11 12 409 13 14 410 On the function of the eyespots in Typophyllum spurioculis 15 16 411 The function of the coxal eye spots of Typophyllum spurioculis are currently unknown. 17 18 19 412 Antipredatory markings have been reported in other orthopteroid insects, for example in the 20 21 413 hind wings of praying (Maldonado, 1970), the wings of (Steiner, 1981) 22 23 24 414 and in the wings of the katydids, Pterochroza ocellata (Castner, 1995). The eyespots, in P. 25 26 415 ocellata (another leaf-mimicking katydid) are usually relatively large, brightly pigmented 27 28 29 416 patches positioned on the lower tegminal surface and hindwings, away from more vital areas 30 31 417 (e.g. the head) (Edmunds, 1974). In response to a physical disturbance, the wings are 32 33 418 suddenly flicked open, startling the predator. It is believed that, when viewed from behind, 34 35 36 419 the spots and abdomen resemble a bird's eyes and beak respectively (Castner, 1995; Nickle 37 38 420 and Castner, 1995). Eyespots reduce predation on conspicuous (non-camouflaged prey), but 39 40 41 421 increase them on camouflaged prey (Stevens et al., 2008). This explains why P. ocellata 42 43 422 eyespots are hidden under the resting tegmina and are only exposed when the insect feels 44 45 46 423 threatened. There is also the benefit of surprise to gain time: fear/flight reaction evoked in a 47 48 424 predator by sudden disconcerting change (Edmunds, 1974). 49 50 51 425 In P. ocellata and other Pterochrozini species (e.g., T. spurioculis) the general morphology is 52 53 426 one of leaf mimicry: they have a general antipredator strategy of crypsis. Crypsis is their 54 55 427 primary defense (Robinson, 1969) and only when this pose is breached by a foraging predator 56 57 58 428 will the secondary defense be brought into play. For P. ocellata this secondary defense is the 59 60 429 large wing eyespots. The secondary strategy of T. spurioculis is to employ its metathoracic 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 19

430 legs and leap forcefully and quickly away in some unpredictable direction in the cluttered 1 2 431 vegetation. The small coxal eyespots of T. spurioculis could have a function other than 3 4 5 432 deterring predators. Larger markings, and more markings are more effective deterrents of 6 7 433 predation than smaller or fewer ones. Eyespots that actually look like eyes work better to 8 9 10 434 deter predators than equally conspicuous markings (De Bona et al., 2015). In T. spurioculis 11 12 435 each spot has a melanised black border, which boosts the contrast between the spot and the 13 14 436 body. These eyespots are localized in the anterior part of the thorax (fore and mid coxae), and 15 16 17 437 close to the head (Fig. 1). Therefore, it seems reasonable to assume they have signal function 18 19 438 between conspecifics. It is possible that the spots have a role in sexual selection. For example, 20 21 22 439 salient spots on cryptic individuals usually increase predation (Stevens et al., 2008), therefore 23 24 440 only the fittest individuals survive to reproduce. The brightness of the spot could be an honest 25 26 27 441 signal of high fitness. However if the spots are adornments influencing female choice then 28 29 442 presumably they should be absent from the female herself, which is not the case in T. 30 31 443 spurioculis (Fig. 1A). 32 33 34 444 The eye pairs of T. spurioculis, which appear on the right and left, are not common as 35 36 445 defensive mechanisms. Eyespots for deterring predators occur usually symmetrically as one 37 38 39 446 ‘eye’ on each side of the body (Hossie, 2014; Janzen et al., 2010). These small ‘paired eye’ 40 41 447 displays of T. spurioculis on the side must be effective laterally, but the reduced size might to 42 43 44 448 be effective against predators approaching laterally. However, it is well documented that male 45 46 449 leaf-mimicking katydids (at least Typophyllum spp.) position at 90 degree plane on the side of 47 48 the much larger female during mating (Braun 2002; Braun, 2015a; Braun, 2015b; 49 450 50 51 451 Montealegre-Z and Morris, 1999) and we observed this behavior in T. spurioculis in caged 52 53 452 specimens). Perhaps the coupling posture is when these eye markings could be most useful, 54 55 56 453 for example for mate recognition. But other Typophyllum species with similar mating 57 58 454 behaviour do not have coxal eyespots. Very little is found in the literature on experimental 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 20

455 data on the eyespot behavioural display in , and this topic deserves further 1 2 456 attention. 3 4 5 457 6 7 458 Acknowledgements 8 9 10 459 This project was developed during the University of Lincoln annual field course in Ecuador, 11 12 460 with thanks to Adrian Godman. We thank all the staff at the Santa Lucía Lodge for hospitality 13 14 15 461 and lab facilities used. The experimental part of this project was funded by the Leverhulme 16 17 462 Trust (grant RPG-2014-284 to FM-Z). We also thank Tom Pike, David Kikuchi, and Thomas 18 19 20 463 Hossie for providing feedback on the discussion of eyespot function. 21 22 464 23 24 25 465 Competing interests 26 27 466 The authors have declared that no competing interests exist. 28 29 30 467 31 32 468 Funding 33 34 35 469 FMZ and TJ are currently sponsored by the Leverhulme Trust (grant No. RPG-2014-284). 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 21

470 References 1 2 3 471 Bailey, W.J., Broughton, W.B., 1970. The mechanics of stridulation in bush crickets 4 472 (Tettigonioidea, Orthoptera) II. Conditions for resonance in tegminal generator. J. Exp. 5 473 Biol. 52, 507-517. 6 474 Bennet-Clark, H.C., 1998. Size and scale effects as constraints in insect sound 7 8 475 communication. Philosophical Transactions of the Royal Society of London B Biological 9 476 Sciences 353, 407-419. 10 477 Bennet-Clark, H.C., 1999a. Resonators in insect sound production: How insects produce 11 478 loud pure-tone songs. Journal of Experimental Biology 202, 3347-3357. 12 13 479 Bennet-Clark, H.C., 1999b. Which Qs to choose: questions of quality in bioacoustics? 14 480 Bioacoustics 9, 351-359. 15 481 Bennet-Clark, H.C., 2003a. Wing resonances in the Australian field cricket Teleogryllus 16 482 oceanicus. J. Exp. Biol. 206, 1479-1496. 17 18 483 Bennet-Clark, H.C., 2003b. Wing resonances in the Australian field cricket Teleogryllus 19 484 oceanicus. Journal of Experimental Biology 206, 1479-1496. 20 485 Braun , H., 2002. Die Laubheuschrecken (Orthoptera, Tettigoniidae) eines 21 486 Bergregenwaldes in Süd-Ecuador, faunistische, bioakustische und ökologische 22 23 487 Untersuchungen, Zoology II. Friedrich-Alexander –Universität, Erlangen-Nürnberg, 24 488 Erlangen, Germany. 25 489 Braun, H., 2015a. The little walking leaves from Iguazú National Park (NE Argentina) – 26 490 comparisons between subtropical and tropical Typophyllum species (Orthoptera: 27 28 491 Tettigoniidae: Pterochrozinae). J. Orthoptera Res. 24, 55-58. 29 492 Braun, H., 2015b. Little walking leaves from southeast Ecuador: biology and taxonomy of 30 493 Typophyllum species (Orthoptera, Tettigoniidae, Pterochrozinae). Zootaxa 4012, 1-32. 31 494 Broughton, W.B., 1964. Function of the mirror in Tettigoniidae Orthoptera. Nature 201, 32 495 949-950. 33 34 496 Castner, J.L., 1995. Defensive Behavior and Display of the Leaf-Mimicking Katydid 35 497 Pterochroza ocellata (L.) (Orthoptera: Tettigoniidae: Pseudophyllinae: Pterochrozini). 36 498 Journal of Orthoptera Research, 89-92. 37 499 Chivers, B.D., Béthoux, O., Sarria-S, F.A., Jonsson, T., Mason, A.C., Montealegre-Z, F., 2017. 38 39 500 Functional morphology of tegmina-based stridulation in the relict species Cyphoderris 40 501 monstrosa (Orthoptera: Ensifera: Prophalangopsidae). Journal of Experimental Biology 41 502 220, 1112-1121. 42 503 Chivers, B.D., Jonsson, T., Jackson, J.C., Kleinhappel, T.K., Shivarova, N., Windmill, J.F.C., 43 44 504 Montealegre-Z, F., 2016. Distribution of sound pressure around a singing cricket: 45 505 radiation pattern and asymmetry in the sound field. Bioacoustics 25, 161-176. 46 506 Cigliano, M.M., Braun, H., Eades, D.C., Otte, D., 2017. Orthoptera Species File. Version 47 507 5.0/5.0. , 2015 ed. The Orthopterists' Society. 48 49 508 De Bona, S., Valkonen, J.K., López-Sepulcre, A., Mappes, J., 2015. Predator mimicry, not 50 509 conspicuousness, explains the efficacy of butterfly eyespots, Proc. R. Soc. B. The Royal 51 510 Society, p. 20150202. 52 511 Edmunds, M., 1974. Defence in animals: a survey of anti-predator defences. Longman 53 54 512 Publishing Group. 55 513 Fletcher, N.H., 1992. Acoustic Systems in Biology. Oxford University Press, Oxford. 56 514 Fletcher, N.H., 2007. Bioacoustics, in: Rossing, T.D. (Ed.), Springer Handbook of 57 515 Acustics, 1st ed. Springer Science+Business Media, LLC New York, pp. 785-804. 58 59 516 Gwynne, D.T., 2001. Katydids and bush-crickets: Reproductive behaviour and evolution 60 517 of the Tettigoniidae. Cornell University Press, Ithaca. 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 22

518 Hemp, C., Heller, K.G., Warchalowska-Sliwa, E., Hemp, A., 2013. The genus Aerotegmina 1 519 (Orthoptera, Tettigoniidae, ): chromosomes, morphological relations, 2 520 phylogeographical patterns and description of a new species. Org Divers Evol 13, 521- 3 4 521 530. 5 522 Hirtenlehner, S., Kung, S., Kainz, F., Romer, H., 2013. Asymmetry in cricket song: female 6 523 preference and proximate mechanism of discrimination. J Exp Biol 216, 2046-2054. 7 524 Hossie, T.J., 2014. An investigation into the ecology and evolution of caterpillar eyespots. 8 9 525 Carleton University Ottawa. 10 526 Janzen, D.H., Hallwachs, W., Burns, J.M., 2010. A tropical horde of counterfeit predator 11 527 eyes. Proceedings of the National Academy of Sciences 107, 11659-11665. 12 528 Jatho, M., Schul, J., Stiedl, O., Kalmring, K., 1994. Specific differences in sound production 13 529 and pattern recognition in tettigoniids. Behav. Processes 31, 293-300. 14 15 530 Jonsson, T., Chivers, B.D., Robson-Brown, K.A., Sarria-S, F.A., Walker, M., Montealegre-Z, 16 531 F., 2017. Chamber music – An unusual Helmholtz resonator for song amplification in a 17 532 Neotropical bush-cricket (Orthoptera, Tettigoniidae). Journal of Experimental Biology 18 533 220, In press. 19 20 534 Leroy, Y., 1966. Signaux acoustiques, comportement et systématique de quelques 21 535 espèces de Gryllides (Orthoptères, Ensifères). Bull. Biol. Fr. Belg. 100, 63-134. 22 536 Maldonado, H., 1970. The deimatic reaction in the praying mantis 23 537 biocellata. Zeitschrift für vergleichende Physiologie 68, 60-71. 24 25 538 Montealegre-Z, F., 2005. Biomechanics of musical stridulation in katydids (Orthoptera: 26 539 Ensifera: Tettigoniidae): an evolutionary approach, Department of Zoology. University of 27 540 Toronto, Toronto, p. 328. 28 541 Montealegre-Z, F., Jonsson, T., Robert, D., 2011. Sound radiation and wing mechanics in 29 30 542 stridulating field crickets (Orthoptera: ). J. Exp. Biol. 214, 2105-2117. 31 543 Montealegre-Z, F., Mason, A.C., 2005. The mechanics of sound production in Panacanthus 32 544 pallicornis (Orthoptera : Tettigoniidae : ): the stridulatory motor 33 545 patterns. J. Exp. Biol. 208, 1219-1237. 34 35 546 Montealegre-Z, F., Morris, G.K., 1999. Songs and Systematics of some Tettigoniidae from 36 547 Colombia and Ecuador, part I. Pseudophyllinae (Orthoptera). J. Orthoptera Res. 8, 163- 37 548 236. 38 549 Montealegre-Z, F., Morris, G.K., 2003. Uchuca giglio-tos, Dectinomima Caudell and their 39 40 550 allies (Orthoptera : Tettigoniidae : Conocephalinae). Trans. Am. Entomol. Soc. 129, 503- 41 551 537. 42 552 Montealegre-Z, F., Morris, G.K., 2004. The spiny devil katydids, Panacanthus Walker 43 553 (Orthoptera : Tettigoniidae): an evolutionary study of acoustic behaviour and 44 45 554 morphological traits. Syst. Entomol. 29, 21-57. 46 555 Montealegre-Z, F., Postles, M., 2010. Resonant Sound Production in Copiphora 47 556 gorgonensis (Tettigoniidae: ), an Endemic Species from Parque Nacional 48 557 Natural Gorgona, Colombia. J. Orthoptera Res. 19, 347-355. 49 50 558 Morris, G.K., 1980. Calling display and mating behaviour of Copiphora rhinoceros Pictet 51 559 (Orthoptera, Tettigoniidae). Anim. Behav. 28, 42-51. 52 560 Morris, G.K., 1999. Song in , in: Davey, K.G. (Ed.), Encyclopedia of 53 561 Reproduction. Academic Press, San Diego, pp. 508–517. 54 562 Morris, G.K., Klimas, D.E., Nickle, D.A., 1989. Acoustic signals and systematics of false-leaf 55 56 563 Katydids from Ecuador (Orthoptera, Tettigoniidae, Pseudophyllinae). Trans Am Entomol 57 564 Soc (Phila) 114, 215-263. 58 565 Morris, G.K., Mason, A.C., 1995. Covert Stridulation - Novel sound generation by a South- 59 566 American Katydid. Naturwissenschaften 82, 96-98. 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 23

567 Morris, G.K., Montealegre-Z, F., 2001. Los Tettigoniidae (Orthoptera: Ensifera) del 1 568 Parque Regional Nacional Ucumarí: Aspectos interesantes de comunicación acústica. 2 569 Rev Colomb Entomol 27, 93-105. 3 4 570 Morris, G.K., Pipher, R.E., 1967. Tegminal amplifiers and spectrum consistencies in 5 571 Conocephalus nigropleurum (Bruner), Tettigoniidae. J. Insect Physiol. 13, 1075-1085. 6 572 Nickle, D.A., 1992. Katydids of Panama (Orthoptera: Tettigoniidae), in: Quintero, D., 7 573 Aiello, A. (Eds.), Insects of Panama and Mesoamerica. Selected Studies. Oxford University 8 9 574 Press, Oxford, pp. 142- 184. 10 575 Nickle, D.A., Castner, J.L., 1995. Strategies utilized by katydids (Orthoptera: 11 576 Tettigoniidae) against diurnal predators in rainforests of northeastern Peru. J. 12 577 Orthoptera Res. 4, 75-88. 13 578 Robillard, T., Montealegre-Z, F., Desutter-Grandcolas, L., Grandcolas, P., Robert, D., 2013. 14 15 579 Mechanisms of high-frequency song generation in brachypterous crickets and the role of 16 580 ghost frequencies. Journal of Experimental Biology 216, 2001-2011. 17 581 Robinson, D., 1990. Acoustic communication between the sexes in bushcrickets, in: 18 582 Bailey W. J. and Broughton, W.B., Rentz, D.C.F. (Eds.), The Tettigoniidae: Biology, 19 20 583 Systematics and Evolution. Crawford House Press, Bathurst, pp. 112-129. 21 584 Robinson, D.J., Hall, M.J., 2002. Sound signalling in Orthoptera. Adv. Insect Physiol. 29, 22 585 151-278. 23 586 Robinson, M.H., 1969. Defenses against visually hunting predators. Evolutionary biology 24 25 587 3, 225-259. 26 588 Sarria-S, F.A., Buxton, K., Jonsson, T., Montealegre-Z, F., 2016. Wing mechanics, 27 589 vibrational and acoustic communication in a new bush-cricket species of the genus 28 590 Copiphora (Orthoptera: Tettigoniidae) from Colombia. Zoologischer Anzeiger 263, 55- 29 30 591 65. 31 592 Sarria-S, F.A., Morris, G.K., Windmill, J.F.C., Jackson, J., Montealegre-Z, F., 2014. Shrinking 32 593 Wings for Ultrasonic Pitch Production: Hyperintense Ultra-Short-Wavelength Calls in a 33 594 New Genus of Neotropical Katydids (Orthoptera: Tettigoniidae). PLoS One 9. 34 35 595 Steiner, A., 1981. Anti-predator strategies. II. Grasshoppers (Orthoptera, Acrididae) 36 596 attacked by Prionyx parkeri and some Tachysphex wasps (Hymenoptera, Sphecinae and 37 597 Larrinae): a descriptive study. Psyche 88, 1-24. 38 598 Stevens, M., Hardman, C.J., Stubbins, C.L., 2008. Conspicuousness, not eye mimicry, 39 40 599 makes “eyespots” effective antipredator signals. Behavioral Ecology 19, 525-531. 41 600 Stumpner, A., Dann, A., Schink, M., Gubert, S., Hugel, S., Ziesmann, J., 2013. True katydids 42 601 (Pseudophyllinae) from Guadeloupe: Acoustic signals and functional considerations of 43 602 song production. Journal of Insect Science 13, 157-157. 44 45 603 Stumpner, A., Meyer, S., 2001. Songs and the Function of Song Elements in Four Duetting 46 604 Bushcricket Species (Ensifera, Phaneropteridae, Barbitistes). J. Insect Behav. 14, 511- 47 605 534. 48 606 Walker, T.J., Dew, D., 1972. Wing movements of calling katydids: fiddling finesse. Science 49 50 607 178, 174-176. 51 608 52 53 609 54 55 56 57 58 59 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 24

610 Figure captions 1 2 3 611 4 5 612 Fig. 1 Typophyllum spurioculis sp. nov. A, habitus of male (lower specimen) and female (top 6 7 613 specimen). Inset: close up view of the female showing the orange eye-spots in the coxo- 8 9 10 614 trochanter membrane (yellow arrows). B-D, pictures illustrating variation in wing 11 12 615 morphology across individuals. 13 14 15 616 16 17 617 Fig. 2 Wing venation pattern. A, line drawings showing variation in wing venation in 10 male 18 19 20 618 left wings. B, the enlarged inset shows all the wings overlaid to highlight the most common 21 22 619 venation pattern. C, the left fore legs of a male and female in antero-posterior view. 23 24 25 620 26 27 621 Fig. 3 Stridulatory file and tooth distribution. A, SEM of the file of an Ecuadorean male. B, 28 29 622 tooth distribution based on inter-tooth distances (filled, black circles) and tooth length 30 31 32 623 variation (open, red circles) across the file. Note that the file clearly exhibits groups of large 33 34 624 teeth with short inter-tooth distances and groups of smaller teeth with large inter-tooth 35 36 37 625 distances. 38 39 626 40 41 42 627 Fig. 4 Description of the calling song based on individuals of two populations. A, B, a single 43 44 628 call composed of two syllables. C, D, high resolution of small segment of the major pulse of 45 46 629 the second syllable, showing the characteristic of two harmonically related carrier 47 48 49 630 frequencies. E, F, spectral analysis of both calls, showing fundamental frequencies at 6.8 and 50 51 631 7.0 kHz, respectively, and a harmonically related peak in each case. 52 53 54 632 55 56 633 Fig. 5 Analysis of the song of an Ecuadorian male T. spurioculis. A, two minute-sequence of 57 58 59 634 the song of one individual showing call groups. B, two consecutive calls extracted from the 60 61 62 63 64 65 Wing resonances in a dead-leaf-mimic katydid 25

635 segment selected by the dashed box. C, close up view of the first call (including only the 1 2 636 closing stroke of the syllable 1, and opening and closing of syllable 2) in B (upper part) and 3 4 5 637 analysis of fundamental frequency variation in time showing that the frequency of calls is 6 7 638 periodically modulated (lower part). 8 9 10 639 11 12 640 Fig. 6 Right mirror resonances in two male T. spurioculis. A, B, close up dorsal view of the 13 14 641 stridulatory fields of two mounted specimens with wings extended. C, D, vibration maps of 15 16 17 642 the mirror of the males shown above. E, F, scanning lattice (dorso-anterior view) of the wings 18 19 643 showing the mirror membrane deflection in 3D. G, right mirror resonance. Black outline: 20 21 22 644 average resonance measured from five males. Shaded areas indicate one standard deviation. 23 24 645 25 26 27 646 Fig. 7 Complete laser scan of both wings showing resonances at different locations. A, male 28 29 647 specimen with the wings extended. Numbers indicate some of the areas of interest. B, same as 30 31 648 A, showing the scanning mesh and density of scanning points over the wings. C, vibration 32 33 34 649 map showing major areas of deflection. D, 3D view of the map in C to show that amplitude of 35 36 650 vibration is dominant in the right mirror. E-H, wing resonances measured from the four areas 37 38 39 651 indicated with numbers 1-4 for each wing in A. 40 41 652 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Table

Table 1: Morphological measurements of Typophyllum spurioculus. F, fore; H, hind; L, left; M, mid; R, right; SD, standard deviation.

Males Females Character SD SD (n=9) (n=2)

Body (incl. tegmen) 36.46 2.23 60.59 4.84

Tegmen 30.35 1.52 50.25 8.48

Pronotum 5.21 0.42 7.25 0.86

F-Femur R&L 9.02 0.63 11.73 1.14

M-Femur R&L 8.89 0.39 11.74 0.03

H-Femur R&L 18.70 1.37 25.61 2.81

F-Tibia R&L 10.14 0.51 12.01 0.67

M-Tibia R&L 9.41 0.45 11.45 0.48

H-Tibia R&L 20.11 1.07 26.34 1.63

Subgenital plate 2.22 0.28 3.79 0.00

Stridulatory file 5.50 0.13 n/a n/a

Fig. 1

A B

♀

C

♂

D

♂ Fig. 2

A B

dorsal anterior

C Female

Male

5 mm 10 mm Fig. 3

A

Anal 2mm Basal

B 0.22 40 0.18 30 0.14 0.1

Inter-tooth 20

distances (µm) 0.06 0 20 40 60 80 100 120 140 160 Anal Tooth number Basal legnth (mm) Tooth Fig. 4

Male from Ucumari, Colombia Male from Sta Lucia, Ecuador A Call B Syllable 1

Opennig pulse 20 ms Closing stroke Syllable 2 C pulse D

0.2 ms E F

0 6.8 7.0

() -5 -10 -15 -20 -25 -30 Relative intensity dB 0 10 20 30 40 50 0 10 20 30 40 50 Frequency (kHz) Frequency (kHz) Fig. 5 Call group

A

60 s

B

2 s

C

2 1 0 -1

Amplitude (Pa) -2 7.5

7.0

6.5

Instantaneous

frequency (kHz)

6.0 0.1 0.12 0.14 0.16 0.18 Time (s) Fig. 6 Male 1 Male 2 A B rostral

caudal C D

EF

G 300 6.6 11.3 13.3 16.6 250

200

150

100

Magnitude (nm/Pa) 50

0

0 5 10 15 20 25 30 35 40 Frequency (kHz) Fig. 7

12 E A 10 LW 8 RW 1

6 3 2 4 4 2 3 4

1 1 2

0 10 9 8 F B 7 2 6 5 4 3 2 1 0 12

10 G 3 C 8

Magnitude (nm/Pa) 6

4

2

0 60 H 50 4 40

D 30

20

10

0 0 5 10 15 20 25 30 35 40 45 50 Frequency (kHz) Supplementary Material Click here to download Supplementary Material: Baker et al. Supplementary Materials.pdf

Author Agreement

We, the authors, wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.

We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed.

We further confirm that the of authors listed in the manuscript has been approved by all of us.

We confirm that we have given due consideration to the protection of intellectual property associated with this work and that there are no impediments to publication, including the timing of publication, with respect to intellectual property. In so doing we confirm that we have followed the regulations of our institutions concerning intellectual property.

The authors:

Andrew Baker

Fabio A. Sarria-S

Glenn K. Morris

Thorin Jonsson

Fernando Montealegre-Z